Parsing Partisanship and Partisan Defection in the Postwar US House.

The singular act of a Member of Congress placing a roll call vote in the US House, either yea or nay, occurred over five and a half million times from the 80th to 100th Congress. Using a dataset that links a large variety of exogenous factors to each of those acts of voting, this paper seeks to identify the development of ‘inter-party partisanship’ (how opposed are the Democrats and Republicans to one another?) and ‘intra-party’ partisanship, (how cohesively do members of the same party vote together?). The first section of this paper examines the scope of roll call voting in the post war US House, detailing the number, distribution, and variety of roll call votes held in that period. Following that, the second section examines methods that can be used to identify inter-party partisanship, including updated uses of Lowell and Rice measures. In addition, party and vote are correlated for each roll call vote to create another measure of inter-party partisanship - one that is explored by Clausen roll call topics. The paper goes on to examine intra-party partisanship by looking at predictors of partisan defection. Using a variety of methods culminating in a comprehensive Logit model, we are able to parse out many of the causal predictors of defection - roll call characteristics, member characteristics, constituency characteristics, and electoral marginality. We find that while no magic bullet exists that can overwhelmingly account for variance in the probability of a member defecting, several patterns emerge that indicate the large role roll call characteristics and member characteristics have in predicting defection. Please email me (after 15 October) at danielsinger@comcast.net for full paper [ABSTRACT FROM AUTHOR]